[K3] Allow tpu to import kimi_k3.common

[majunze2001]

Purpose

TPU platform plugins reuse Kimi K3's shared multimodal preprocessing while
registering their own model implementations. Avoid eagerly importing a GPU
implementation when the active platform's device type is TPU.

CUDA and ROCm model selection remains unchanged.

Not a duplicate

PR #51196 disables dynamic torch.compile for the Kimi vision encoder on TPU;
it does not address eager GPU model imports from the Kimi K3 package. Searches
for open PRs covering Kimi K3 TPU package imports found no duplicate.

Testing

uvx --from 'pre-commit>=4.5.1' pre-commit run --files vllm/models/kimi_k3/__init__.py
All applicable hooks passed.

git diff --check
Passed.

A TPU import smoke test confirmed that shared Kimi K3 preprocessing can be
imported without loading the NVIDIA or AMD model packages.

AI assistance

AI assistance was used to review the change and prepare this description. I
reviewed every changed line and can explain and defend the change end-to-end.

[claude]

Claude Code Review

This pull request is from a fork — automated review is disabled. A repository maintainer can comment @claude review to run a one-time review.

[ZJY0516]

/ci run

[github-actions]

✅ Triggered Buildkite CI #83045 for commit 00cb767457b7.